Spectroscopy and photophysics of trimethyl-substituted derivatives of 5-deazaalloxazine. Experimental and theoretical approaches

Article abstract of DOI:10.1016/j.molstruc.2014.09.028

Photophysical and structural properties of 5-deazaalloxazine derivatives were studied, with methyl substituent at the position 8 or 9 of the benzene ring and at the N(1) and N(3) positions. Crystal structures of the compounds were determined by X-ray diff

Full text of DOI:10.1016/j.molstruc.2014.09.028

Journal of Molecular Structure 1079 (2015) 139–146
Contents lists available at ScienceDirect
Journal of Molecular Structure
journal homepage: www.elsevier.com/locate/molstruc
Spectroscopy and photophysics of trimethyl-substituted derivatives
of 5-deazaalloxazine. Experimental and theoretical approaches
a
a
a
b
Dorota Prukała a, , Mateusz Gierszewski , Maciej Kubicki , Tomasz Pe˛dzinski , Ewa Sikorska ,
⇑
´
Marek Sikorski a
a
´
´
Faculty of Chemistry, Adam Mickiewicz University in Poznan, Umultowska 89B, 61-614 Poznan, Poland
b Faculty of Commodity Science, Poznan´ University of Economics, al. Niepodległo´sci 10, 61-875 Poznan´, Poland
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
ꢀ
Photophysical properties of newly
synthesized 5-deazaalloxazines were
presented.
ꢀ
ꢀ
Fluorescence lifetime measurements
were performed in different solvents.
The crystal structures of investigated
compounds were determined by X-
ray analysis.
ꢀ
ꢀ
S0 ? Si, S0 ? Ti and T1 ? Ti transitions
were calculated, using the TD-DFT
methods.
Electronic structures of 5-
deazaalloxazines were calculated by
the same methods.
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 4 May 2014
Received in revised form 6 September 2014
Accepted 11 September 2014
Available online 21 September 2014
Photophysical and structural properties of 5-deazaalloxazine derivatives were studied, with methyl sub-
stituent at the position 8 or 9 of the benzene ring and at the N(1) and N(3) positions. Crystal structures of
the compounds were determined by X-ray diffraction method.
Electronic structure and S0 ? Si, S0 ? Ti and T1 ? Ti transition energies and oscillator strengths were cal-
culated for 1,3,8- and 1,3,9-trimethyl-5-deazaalloxazine, using the TD-DFT methods. The calculations sug-
gest that for 5-deazaalloxazine derivatives the lowest-lying singlet excited states have pure
This is in contrast to the alloxazine derivatives, where two lowest-lying transitions are of
p
,
p⁄ nature.
,
p⁄ and n,p⁄
Keywords:
Alloxazines
p
character, being isoenergetic in most cases. The theoretical data are compared to the experimental results,
involving steady-state and time-resolved spectra. 1,3,8-Trimethyl-5-deazaalloxazine and 1,3,9-trimethyl-
5-deazaalloxazine have longer fluorescence lifetimes and higher values of fluorescence quantum yield
compared to the alloxazine derivatives in the same solvents, whereas the absorption maxima and emis-
sion maxima of studied compounds are blue-shifted as compared to trimethyl derivatives of alloxazine.
Ó 2014 Elsevier B.V. All rights reserved.
5-Deazaalloxazines
DFT study
Photophysics
UV–Vis spectroscopy
X-ray crystallography
Introduction
mechanistic studies with numerous flavoproteins [2]. For the past
three
decades,
5-deazaflavins
(pyrimido[4,5-b]quinolin-
5-Deazaflavin (with tricyclic 5-deazaisoalloxazine ring) occurs
as a natural cofactor in yellow chromophores [1] in various bacte-
ria and has been extensively used as an artificial coenzyme in
2,4(3H,10H)-diones), where N-5 of the flavin is replaced by –CH,
have received much attention, because of the discovery that they
serve as cofactors for several flavin-dependent enzymatic reactions
[3]. 5-Deazaflavins are potential riboflavin antagonists with their
own redox system, different from that of riboflavin [3]. 5-Deazaal-
loxazines (pyrimido[4,5-b]quinoline-2,4(1H,3H)-diones) are ana-
⇑
Corresponding author. Tel.: +48 61 8291593; fax: +48 61 8291555.
E-mail address: dprukala@amu.edu.pl (D. Prukała).
http://dx.doi.org/10.1016/j.molstruc.2014.09.028
0022-2860/Ó 2014 Elsevier B.V. All rights reserved.

140
D. Prukała et al. / Journal of Molecular Structure 1079 (2015) 139–146
logs of 5-deazaflavins, they are also analogs of alloxazines
(benzo[g]pteridine-2,4(1H, 3H)-diones).
Experimental
Fig. 1 presents the structures of investigated compounds with
structures of 5-deazaflavin and alloxazine for comparison.
It is postulated that alloxazines and 5-deazaalloxazines, for
example lumichrome (7,8-dimethylalloxazine) and 5-deazalumi-
chrome (7,8-dimethyl-5-deazaalloxazine) have similar properties
in their ground and excited singlet state, namely solvatochromism,
ground and excited state pKa, and small differences between their
ground and excited state dipole moments [4]. Additionally, the
common structure of 5-deazaflavin and 5-deazaalloxazine is that
both contain the pyrimido[4,5-b]quinoline moiety and this type
of compound is of importance owing to its various biological activ-
ities, especially antimalarial, antibacterial, analgesic and antitumor
activities [3,5–9], which have led to its use in industrial drugs [10–
12,5]. Different derivatives of 5-deazaalloxazine are also consid-
ered as suicide substrates for adenosine deaminase and inosine
5-monophosphate dehydrogenase [13].
Recently, there is growing interest in development of fluores-
cent nucleosides based on 5-deazaalloxazine that maintain the
hydrogen-bonding properties of natural nucleoside bases [14,15].
To our surprise, although derivatives of 5-deazaalloxazine are of
great interest in many aspects, there is only few literature data
regarding their photophysical properties [16–18]. So the aim of
the present work is to characterize spectroscopic and photophysi-
cal properties of 1,3,8-trimethyl-5-deazaalloxazine (138Me-5DAll)
and 1,3,9-trimethyl-5-deazaalloxazine (139Me-5DAll). The experi-
mental data are compared with theoretical studies by using time-
dependent density functional theory (TD-DFT). Accordingly to our
recent results [18], 5-deazaalloxazines show high values of fluores-
cence quantum yields, and show longer fluorescence lifetimes than
their alloxazine analogues [19–23]. In addition, the data from the-
oretical calculations indicated that the lowest-energy transitions of
Materials
Solvents: methanol (MeOH), ethanol (EtOH), 2-propanol (2-
PrOH), acetonitrile (ACN), chloroform (CHCl3), methylene chloride
(CH2Cl2), 1,4-dioxane (Diox), were of spectroscopic or HPLC grade
(Aldrich, Merck). They were dried before use with 3 Å or 4 Å molec-
ular sieves from Fluka.
General procedure for the preparation of pyrimido[4,5-b]quinoline-
2,4(1H,3H)-diones Step 1 [27]. To a stirred solution of o-methylani-
line (or m-methylaniline) (3 ml) was added hydrochloric acid (36%,
6 mmol) and 1,3-dimethyl-6-aminouracil (6 mmol). The mixture
was heated at 170–180 °C for 4–5 h. Diethyl ether was added after
cooling and precipitated solids were collected by filtration. The
corresponding 6-anilinouracils were obtained in 40–98% yields.
Step 2. The appropriate 6-anilinouracil derivative (1 g) was
mixed with DMF (2 ml) and heated to 40 °C under stirring. Then
phosphorus oxychloride (2.5 ml) was dropped into the mixture.
The mixture was than heated at 80 °C for 0.5 h. After cooling,
excess DMF and POCl3 were evaporated in vacuo. The solids were
diluted with 5% NH4OH and then collected by suction filtration
and washed with water. Crude products were recrystallized from
glacial acetic acid. Obtained crystals were suitable for X-ray
analysis.
Purity of all of the compounds studied was checked by thin
layer chromatography and elemental analysis.
1,3,8-Trimethylpyrimido[4,5-b]quinoline-2,4(1H,3H)-dione,
(138Me-5DAll) (isolated yield 55%; m.p. 230 °C); 1H NMR (TFA-d): d
ppm: 2.79 (s, 3H, H3C-C8); 4.10 (s, 3H, H3C-N1); 3.71 (s, 3H, H3C-
N3); 7.88 (d, 1H, H-7); 8.16 (s, 1H, H-9); 8.29 (d, 1H, H-6); 9.77
(s, 1H, H-C5); Anal. Calcd for C14H13N3O2: C, 65.88; H, 5.10; N,
16.47. Found: C, 66.01; H, 4.99; N, 16.38.
1,3,9-Trimethylpyrimido[4,5-b]quinoline-2,4(1H,3H)-dione,
(139Me-5DAll) (isolated yield (54%; m.p. 214 °C); 1H NMR (TFA-d):
d ppm: 2.94 (s, 3H, H3C-C9); 4.16 (s, 3H, H3C-N1); 3.71 (s, 3H, H3C-
N3); 7.98 (t, 1H, H-7); 8.23 (d, 1H, H-8); 8.30 (d, 1H, H-6); 9.87 (s,
1H, C-5); Anal. Calcd for C14H13N3O2: C, 65.88; H, 5.10; N, 16.47.
Found: C, 65.92; H, 5.01; N, 16.40.
5-deazalloxazines have pure
derivatives of alloxazine the lowest
p
,
p⁄ character, as in the case for
p⁄ transition is accompanied
p
,
by closely-located n,p⁄ transition [19–24]. 138Me-5DAll and
139Me-5DAll were synthesized using a method described by Yon-
eda et al. [25,26] in a two-step synthesis. We also present the
detailed crystal structures of the investigated compounds.
Fig. 1. Structures of 1,3,8-trimethyl-5-deazaalloxazine (138Me-5DAll) and 1,3,9-trimethyl-5-deazaalloxazine (139Me-5DAll), and their analogues: 5-deazaflavin and
alloxazine.

D. Prukała et al. / Journal of Molecular Structure 1079 (2015) 139–146
141
Methods
Crystallographic data (excluding the structure factors) for the
structural analysis have been deposited with the Cambridge Crys-
tallographic Data Centre, Nos. CCDC-770192 (138Me-5DAll), and
CCDC-770191 (139Me-5DAll). Copies of this information may be
obtained free of charge from: The Director, CCDC, 12 Union Road,
Cambridge, CB2 1EZ, UK. Fax:+44(1223)336 033, e-mail: depos-
it@ccdc.cam.ac.uk, or www: www.ccdc.cam.ac.uk.
UV/Vis measurements
All solutions were prepared on the same day as their absor-
bance, steady-state fluorescence, fluorescence excitation spectra
were recorded and time-resolved fluorescence measurements per-
formed. As the maximum of absorption, kmax, for the lowest-energy
transitions did not vary much between solvents, it was determined
with three replicates at a reduced scanning rate. The same precau-
tions were taken for the emission kmax, the latter was also deter-
mined for several kexc values.
UV–Vis absorption spectra were recorded on JASCO V-650 spec-
trophotometer and on a Jobin Yvon-Spex Fluorolog 3-22 spectro-
fluorometer, using the option to measure absorption spectra;
steady-state fluorescence excitation and emission spectra were
also recorded on the same spectrofluorometer.
Fluorescence quantum yield were calculated relative to lumi-
chrome as standard (UF = 0.028 in acetonitrile [20]). We used the
method described by Horiba Jobin–Yvon (gradient method), with
simultaneous measurements of absorption and fluorescence on
the same machine (www.jyhoriba.co.uk). Maximum absorption
of all solutions was kept below 0.1. The estimated absolute uncer-
tainty of the calculated fluorescence quantum yield in 20 %.
TD-DFT calculations
Theoretical calculations have been studied by means of time-
dependent density-functional theory (TD-DFT). The TD-DFT calcu-
lations were calculated using the B3LYP hybrid method [32] in con-
junction with a modest 6-31 + g(d) split-valence polarized basis set
[33]. Excitation energies and transition intensities were calculated
for the optimized ground-state geometries. Oscillator strengths
were determined in the dipole length representation. Calculations
were performed using the GAUSSIAN 09 package of ab initio pro-
grams [34].
Results and discussion
Steady-state spectra of 138Me-5DAll and 139Me-5DAll
Absorption properties
Life-time measurements
Absorption spectra of 138Me-5DAll and 139Me-5DAll in
selected organic solvents are presented in Fig. 2A and B. Table 1
summarizes absorption parameters for 138Me-5DAll and 139Me-
5DAll. The compounds have two absorption maxima in the low-
est-energy part of the spectrum (k1 and k2), usually well separated
All fluorescence lifetime measurements were performed using a
time-correlated single-photon counting (TCSPC) method. Decays
were measured with an IBH Consultants (Glasgow, Scotland) Sys-
tem 5000 fluorescence lifetime spectrometer equipped with a Nan-
oLED diode (kexc = 374 nm, fwhm ꢁ 800 ps) and with a Pico-Quant
(Berlin, Germany) FluoTime 300 spectrometer with a 377 nm
pulsed LED, fwhm ꢁ 500 ps) as an excitation source. Both instru-
ments are capable of measuring fluorescence lifetimes of 400 ps
or shorter. Deconvolution of the fluorescence decay curves was
performed using Version 4 of the IBH Consultants software (for
374 nm excitation) or FluoFit Fluorescence Decay Data Analysis
Software (version 4.55) from PicoQuant (for 377 nm excitation).
and with some structure, corresponding to two independent
transitions.
p,
p⁄
As we can see in Fig. 2A and B and Table 1, the two lowest-
energy absorption maxima for 138Me-5DAll and 139Me-5DAll
are located about 360 nm (k1) and about 315 nm (k2), and their
exact positions practically are not dependent on solvent. The posi-
tion of maximum of the lowest-energy band (k1) depends on the
location of the methyl substituent on the benzene ring. Comparing
the spectral properties of investigated compounds with those of 8-
methyl-5-deazaalloxazine and 9-methyl-5-deazaalloxazine (which
are unsubstituted on nitrogen atoms) [18], we note a small shift in
the positions of the band maxima, and a change in the band shapes
of the absorption spectra. In aprotic solvents, the k1 maxima of the
lowest-energy bands in 8-methyl- and 9-methyl-5-deazaallox-
azine are shifted hipsochromically comparing to the corresponding
maxima of 138Me-5DAll and 139Me-5DAll, by about 6 nm.
Additionally, location of the two absorption maxima of 138Me-
5DAll and 139Me-5DAll, are generally hipsochromically shifted, as
compared to their ‘‘aza’’ analogues, namely 8-methyl- and 9-meth-
ylalloxazines [24]. The lowest-energy absorption bands of 5-deaza-
alloxazines, where the N(5) atom in the alloxazine ring is replaced
by the C atom, have a more pronounced vibrational structure than
the respective alloxazine absorption bands.
1H NMR measurements
1H NMR spectra for 138Me-5DAll and 139Me-5DAll were
recorded in TFA-d, as their solubility in typically used DMSO-d6
was insufficient. The 1H NMR spectra were recorded on a Varian
Gemini 300 (300 MHz) Spectrometer. The internal standard was
TMS. 1H NMR spectra of compounds studied were analyzed by
comparison to spectra calculated using the ACD/H NMR predictor
[28].
X-ray diffraction analysis
Diffraction data for 138Me-5DAll and for 139Me-5DAll were
collected at 100 K on an Agilent Technologies XCALIBUR diffrac-
tometer with EOS CCD detector using graphite-monochromated
Mo Ka radiation (k = 0.71073 Å). The data were collected using
the -scan technique to a maximum 2h value of 60° and corrected
x
for Lorentz and polarization effects [29]. Accurate unit-cell param-
eters were determined by a least-squares fit of 1440 (138Me-
5DAll), and 14,443 (139Me-5DAll) reflections of the highest inten-
sity, chosen from the whole experiment. The calculations were
mainly performed within the WinGX program system [30]. The
structure was solved by direct methods with SIR92 [31] refined
by the full matrix least-squares method with SHELXL97. Non-
hydrogen atoms were refined anisotropically. The hydrogen atoms
from methyl groups were placed geometrically in idealized posi-
tions, and refined as rigid groups with their Uiso’s as 1.5 times
Ueq of the appropriate carrier atom; all other hydrogen atoms were
found in subsequent difference Fourier maps and isotropically
refined.
Emission properties
In Fig. 2C and D, we show the emission spectra of 138Me-5DAll
and 139Me-5DAll in different organic solvents. Table 2 presents
the emission parameters including the emission maxima, the fluo-
rescence quantum yields, the fluorescence lifetime, the radiative
rate constants, and the sum of the non-radiative rate constants of
138Me-5DAll and 139Me-5DAll. The absorption and the corrected
fluorescence excitation spectra of all of the compounds agree well
with each other.
Typical emission spectra of investigated compounds show a sin-
gle bands of Gaussian shape. The emission maxima appear at about
420 nm (for 139Me-5DAll). But emission bands for 138Me-5DAll

142
D. Prukała et al. / Journal of Molecular Structure 1079 (2015) 139–146
B
10000
A
10000
8000
6000
4000
2000
0
CHCl3
8000
CHCl3
CH2Cl2
ACN
Cl
CH2
2
ACN
6000
4000
2000
0
MeOH
EtOH
2-PrOH
Diox
MeOH
EtOH
2-PrOH
Diox
300
350
400
450
300
350
400
450
λ/ nm
λ / nm
D
C
CHCl3
AC2N
CHCl3
CH2Cl2
ACN
CH Cl2
MeOH
EtOH
2-PrOH
Diox
MeOH
EtOH
2-PrOH
Diox
400
450
500
550
600
400
450
500
550
600
λ / nm
λ / nm
Fig. 2. A – Absorption spectra of 138Me-5DAll, B – absorption spectra of 139Me-5DAll, C – emission spectra (kexc = 350 nm) of 138Me-5DAll, D – emission spectra
(kexc = 360 nm) of 139Me-5DAll, in selected organic solvents.
Table 2
Table 1
Photophysical data for the singlet states for 138Me-5DAll and 139Me-5DAll.
Absorption parameters k/nm for 138Me-5DAll and 139Me-5DAll.
kF/nma
UF
s
F/nsa
kr/108 sꢂ1a
R
knr/108 sꢂ1a
a
Solvent
Solvent
k1/nma
k2/nma
138Me-5DAll
CHCl3
CH2Cl2
ACN
2-PrOH
EtOH
MeOH
138Me-5DAll
CHCl3
CH2Cl2
ACN
2-PrOH
EtOH
MeOH
406
405
405
408
409
409
0.11
0.02
0.14
0.15
0.15
0.14
2.21
2.01
2.45
2.96
3.03
2.82
0.50
0.09
0.57
0.51
0.49
0.50
4.01
4.88
3.51
2.87
2.81
3.05
355
355
353
353
353
352
318
318
314
317
316
316
139Me-5DAll
CHCl3
CH2Cl2
ACN
2-PrOH
EtOH
MeOH
139Me-5DAll
CHCl3
CH2Cl2
ACN
2-PrOH
EtOH
MeOH
417
417
417
421
423
425
0.17
0.04
0.21
0.27
0.26
0.28
3.92
3.87
4.51
6.89
7.27
7.90
0.43
0.10
0.47
0.39
0.36
0.35
2.11
2.48
1.75
1.06
1.02
0.91
362
362
359
360
359
361
318
317
313
316
315
315
a
kF the fluorescence emission maximum, UF the fluorescence quantum yield, sF
the fluorescence lifetime, kr the radiative rate constant and Rknr the sum of non-
a
k1, k2 are the positions of the two lowest-energy bands in the absorption
spectra.
radiative rate constants.
Photophysics of 138Me-5DAll and 139Me-5DAll
are blue-shifted if compared to the emission bands of 139Me-5DAll
in the same solvents. The emission maxima of 138Me-5DAll are
located at about 407 nm. The solvent polarity affects the exact
position of the emission maxima; however this effect is rather
small, particularly for 138Me-5DAll. There is a red shift of emission
maxima in polar solvents (alcohols) as compared to apolar and
aprotic solvents for both compounds. Interestingly, the substitu-
tion of the N(1) and N(3) by methyl groups does not change this
tendency, thus the emission maxima for the compounds 138Me-
5DAll and 139Me-5DAll are almost the same as those for their
unsubstituted analogues: 8-methyl-5-deazaalloxazine and 9-
methyl-5-deazalloxazine [18]. Generally, the fluorescence emis-
sion maxima of alloxazines are red shifted as compared to those
of their 5-deazaalloxazine analogues in the same solvents, as obvi-
ous from the data in literature [24].
The fluorescence quantum yields, the lifetimes, the radiative
and sum of the non-radiative decay constants for the lowest
excited singlet state of 138Me-5DAll and 139Me-5DAll are pre-
sented in Table 2. The radiative and non-radiative decay constants
were calculated as presented below:
X
kr ¼ UF
=
sF
;
knr ¼ ð1 ꢂ UFÞ=sF
:
In all cases the fluorescence decays are well modelled by single
exponential functions, as shown by the usual statistical criteria of
‘‘goodness-of-fit’’.
Generally, the fluorescence quantum yields are high, the exact
values, depending on solvent and compound, and are between
0.02 and 0.27. The higher values of fluorescence quantum yield

D. Prukała et al. / Journal of Molecular Structure 1079 (2015) 139–146
143
were noted in protic solvents (alcohols) compared to other sol-
vents for the same derivative. 138Me-5DAll shows lower values
of fluorescence quantum yield if comparing those of 139Me-5DAll
at the same conditions. The quantum yields of 138Me-5DAll and
139Me-5DAll are about ten-fold higher as reported for their allox-
azine analogues (8-methyl-5-deazaalloxazine and 9-methyl-5-
deazalloxazine) [18] in the same solvents. The fluorescence life-
times (sF) of 138Me-5DAll are shorter than those of 139Me-5DAll
and additionally are longer than for their ‘‘aza’’ analogous. This
shows that the location of the methyl group on the alloxazinic skel-
eton affects the fluorescence lifetime (see Table 2 for details). Both
of the investigated 5-deazaalloxazines have longer fluorescence
lifetimes in alcohols than those in aprotic solvents. The radiative
and non-radiative decay constants for both 5-deazaalloxazine
derivatives (Table 2) show that the decay of their singlet state is
dominated by the rates of non-radiative processes and are about
two-ten-fold higher than the radiative constants.
Crystal structures
X-ray diffraction analysis shows that crystals of 138Me-5DAll
(0.4 ꢃ 0.4 ꢃ 0.1 mm) and 139Me-5DAll (0.3 ꢃ 0.2 ꢃ 0.15 mm) have
similar structures. The molecules of both compounds are almost
planar and are connected by approximately linear hydrogen bonds
in the crystals. Crystal structure and structure refinement parame-
ters of 138Me-5DAll and 139Me-5DAll are given in Table 3. Fig. 3
presents an ORTEP drawing of 138Me-5DAll and 139Me-5DAll
with the numbering scheme of the individual atoms, used also in
the next part of the paper.
The molecules of 138Me-5DAll and 139Me-5DAll are arranged
into molecular tapes, with the molecules in these tapes connected
by the relatively strong intermolecular C–Hꢄ ꢄ ꢄO hydrogen bonds,
forming oligomers. The C6–H6ꢄ ꢄ ꢄO4 and C5–H5ꢄ ꢄ ꢄO4 hydrogen
bonds in both 138Me-5DAll and 139Me-5DAll connect molecules
into centrosymmetric dimmers. These dimmers are further con-
nected to other dimmers and make infinite tapes by means of
Fig. 3. Anisotropic-ellipsoid representation of 138Me-5DAll (top) and 139Me-5DAll
(bottom), together with the numbering scheme. The ellipsoids are drawn at the 50%
probability level; hydrogen atoms are represented by spheres of arbitrary radii.
C7–H7ꢄ ꢄ ꢄO2 bonds. The hydrogen-bonded molecular tapes in the
crystal structure of 138Me-5DAll and 139Me-5DAll are shown in
Fig. 4. The tapes of molecules are stacked one onto another with
an interplanar distance of ca. 3.4 Å.
The data on the hydrogen bonds are shown in Table 4.
Table 3
Crystal structures and structure refinement parameters for 138Me-5DAll and 139Me-
5DAll.
Singlet and triplet states: theoretical approach
Compounds
138Me-5DAll
14H13N3O2
255.27
Triclinic
P ꢂ 1
139Me-5DAll
Chemical formula
Formula weight
Crystal system
Space group
C
C14H13N3O2
255.27
Triclinic
P ꢂ 1
We employ the quantum-chemical calculations to determine
the lowest energy singlet–singlet electronic transitions (S0 ? Si)
as well as spin-forbidden S0 ? Ti transitions from the optimized
ground state geometry, whereas T1 ? Ti excitation energies and
transition intensities were determined for the optimized geometry
of the lowest triplet state (T1), using the unrestricted method suit-
able for open shell systems. We compare theoretical data obtained
at the TD-DFT rB3LYP/6-31 + G(d) methods with experimental sin-
glet–singlet absorption spectra. Theoretically predicted values
requiring a shift towards the red for about 600–1000 cmꢂ1 to
reproduce experimental spectra.
Fig. 5 shows the predicted lowest-energy singlet–singlet transi-
tions of 138Me-5DAll (A) and 139Me-5DAll (B) together with
experimental spectra of absorption in 1,4-dioxane. The lowest cal-
culated energies of singlet–singlet transitions for both compounds
were included in Table 5 whereas the lowest predicted energies of
singlet–triplet transitions were presented in Table 6.
Unit dimensions
a (Å)
b (Å)
c (Å)
6.1163(7)
7.1376(7)
15.2298(11)
95.414(7)
100.341(8)
111.084(10)
601.06(10)
2
7.9840(9)
8.7710(10)
9.1420(11)
81.712(8)
73.453(8)
82.701(9)
604.77(12)
2
a
(°)
b (°)
c
(°)
Volume (Å3)
Z
Calculated density (g cmꢂ3
)
1.41
268
0.097
1.40
268
0.097
F(000)
Absorption coefficient (mmꢂ1
)
H
range for data collection (°)
3.11–25.00
ꢂ7 6 h 6 7
ꢂ8 6 k 6 7
ꢂ13 6 l 6 18
3085
2084 (0.014)
1226
0.050,0.135
0.085, 0.143
1.017
3.40–27.00
ꢂ10 6 h 6 10
ꢂ11 6 k 6 11
ꢂ11 6 l 6 11
14,735
2632 (0.018)
2236
0.046, 0.133
0.052, 0.138
1.055
Limits of hkl
Reflections: collected
Unique (Rint
According to data presented in Fig. 5 and Table 5, the calculated
)
lowest-energy transitions S0 ? S1 are of pure
p ?
p⁄ character at
For I > 2 (I)
r
approximately 29,300 cmꢂ1 for 138Me-5DAll and 28,500 cmꢂ1 for
139Me-5DAll.
R(F), wR(F2) for reflections with I > 2
R(F), wR(F2) for all reflections
Goodness-of-fit on F2
r
(I)
The observed lowest-energy absorption bands for 138Me-5DAll
Max/min
D
q
(e Åꢂ3) in final
D
F map
0.23/ꢂ0.22
0.19/ꢂ028
and 139Me-5DAll can be attributed to the
p ?
p⁄ transitions and

144
D. Prukała et al. / Journal of Molecular Structure 1079 (2015) 139–146
Wavelength / nm
300 350
250
400
450 500
A
0.8
0.6
0.4
0.2
138Me-5DAll
0.0
40000
30000
20000
10000
0
40000
35000
30000
25000
20000
Wavenumber / cm-1
Wavelength / nm
250
300
350
400
450 500
B
0.8
0.6
0.4
0.2
139Me-5DAll
0.0
40000
30000
20000
10000
0
40000
35000
30000
25000
-1
20000
Wavenumber / cm
Fig. 5. A – Absorption spectra of 138Me-5DAll; B – Absorption spectra of 139Me-
5DAll, both 1,4-dioxan. Predicted transition energies and oscillator strengths f are
indicated by solid vertical lines. The (prohibited) transitions involving the n,p⁄
singlet states are marked by red triangles. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 4. Hydrogen-bonded molecular tape in the crystal structures of 138Me-5DAll
(A) and 139Me-5-DAll (B). Hydrogen bonds are depicted by dashed lines.
Table 5
The lowest predicted [rB3LYP/6-31 + G(d)] singlet excitation energies for isolated
molecule, starting from the ground state of 138Me-5DAll and 139Me-5DAll with the
corresponding oscillator strengths, f.
Table 4
Hydrogen bonds with their lengths (in Å) and angles (in °). Numbering of atoms agree
with Fig. 3.
S0 ? Si
138Me-5DAll
139Me-5DAll
D
H
A
D–H
Hꢄ ꢄ ꢄA
Dꢄ ꢄ ꢄA
D–Hꢄ ꢄ ꢄA
E ꢃ 10ꢂ3/cmꢂ1
f
E ꢃ 10ꢂ3/cmꢂ1
f
138Me-5DAll
?S1
?S2
?S3
?S4
?S5
?S6
?S7
?S8
?S9
?S10
29.3
32.3a
32.7
37.4a
38.2
40.2
41.0a
41.1
41.9a
43.8
0.0590
0.0004
0.1485
0.0001
0.0626
0.1287
0
0.7931
0
0.4323
28.5
32.2a
32.6
37.1a
38.0
39.9
41.0a
41.2
41.8a
43.9
0.0487
0.0003
0.1063
0.0001
0.0982
0.4431
0
0.3759
0
0.3838
C-5
C-6
C-7
H-5
H-6
H-7
O-4
O-4
O2
0.96(2)
0.95(2)
0.96(2)
2.40(2)
2.59(2)
2.36(2)
3.286(3)
3.363(3)
3.275(3)
153.2(15)
139.2(18)
159.0(18)
139Me-5DAll
C-5
C-6
C-7
H-5
H-6
H-7
O-4
O-4
O-2
0.962(16)
0.979(16)
0.970(18)
2.580(16)
2.580(16)
2.524(18)
3.4653(15)
3.4499(17)
3.3316(16)
151.4(12)
148.1(12)
140.7(14)
a
n,p⁄ transition, otherwise p⁄
p, .
compared to maxima of absorption spectra registered in 1,4-
dioxan, see Fig. 5.
It is noteworthy, that the lowest-energy transitions for 1,3,8-
trimethylalloxazine and 1,3,9-trimethylalloxazine, the two com-
pounds which are analogues of investigated in this study com-
accompanied with n,p⁄ transitions. The energy gap between the
two lowest singlet excited states is |E ⁄ ꢂ E
| = 3400 cmꢂ1 for
p
,
p
p,p⁄
pounds, have
p
,
p⁄ character, accompanied with closely located
138Me-5DAll. The energy separation of the two first singlet states
n,p⁄ transitions [24]. In consequence, 138Me-5DAll and 139Me-
5DAll have higher values of fluorescence quantum yields than their
‘‘aza’’ analogues. The S0 ? S2 transitions for both investigated com-
for 139Me-5DAll is |E ⁄ ꢂ E
| = 4100 cmꢂ1
.
p
,p
p,p⁄
Fig. 6 depicts the shape of the highest occupied molecular orbi-
tals (HOMO) and the lowest unoccupied molecular orbitals
(LUMO), involved in the transitions to the low-lying excited states
pounds have dominant
p,
p⁄ character (Table 5), but they are

D. Prukała et al. / Journal of Molecular Structure 1079 (2015) 139–146
145
Table 6
Table 7
The lowest predicted [rB3LYP/6-31 + G(d)] S0 ? Ti excitation energies of 138Me-5DAll
and 139Me-5DAll with their corresponding oscillator strengths, f.
The lowest predicted [uB3LYP/6-31 + G(d)] T1 ? Ti excitation energies of 138Me-
5DAll and 139Me-5DAll with their corresponding oscillator strengths, f. (The
detectable, theoretically predicted transitions are marked as bold numbers).
S
0 ? Ti
138Me-5DAll
139Me-5DAll
T
1 ? Ti
138Me-5DAll
139Me-5DAll
E ꢃ 10ꢂ3/cmꢂ1
f
E ꢃ 10ꢂ3/cmꢂ1
f
E ꢃ 10ꢂ3/cmꢂ1
f
E ꢃ 10ꢂ3/cmꢂ1
f
?T1
?T2
?T3
?T4
?T5
21.9
25.1
30.2
30.7
33.2
0
0
0
0
0
20.9
25.4
30.0
30.2
32.9
0
0
0
0
0
?T2
?T3
?T4
?T5
?T6
?T7
?T8
?T9
?T10
?T11
?T12
?T13
?T14
?T15
?T16
6.9
11.6
12.9
15.6
16.3
16.4
18.3
21.1
22.7
25.8
26.7
29.2
29.7
29.9
31.4
0.0038
0.0001
0.0066
0.0049
0.0019
0.0001
0.0133
0
0.2064
0.0184
0.0004
0.0004
0.0034
0.0004
0.3127
7.3
11.4
12.8
15.1
15.9
16.1
19.1
20.7
22.8
25.6
26.9
29.3
29.8
30.7
31.9
0.0069
0.0001
0.0045
0.0015
0.0022
0.0007
0.0249
0
0.1994
0.0208
0
0.0008
0.0045
0
0.2781
Table 8
Bond lengths (in Å) and dipole moment in S0, S1 and T1 states for 138Me-5DAll and
139Me-5DAll.
Bond
138Me-5DAll
S0 S1
139Me-5DAll
S0 S1
T1
T1
= 3.3 D) (l = 5.8 D)
(l
= 4.2 D) (
l
= 4.1 D) (
l
= 5.9 D) (
l
= 3.5 D) (
l
N10–C10A 1.32
C10A–C4A 1.43
1.34
1.43
1.41
1.37
1.47
1.44
1.23
1.36
1.47
1.44
1.24
1.43
1.37
1.39
1.43
1.38
1.47
1.40
1.23
1.39
1.47
1.41
1.23
1.45
1.32
1.44
1.38
1.39
1.47
1.39
1.22
1.40
1.47
1.40
1.23
1.47
1.34
1.43
1.41
1.37
1.47
1.43
1.24
1.36
1.47
1.44
1.24
1.44
1.37
1.39
1.43
1.38
1.47
1.40
1.23
1.39
1.47
1.42
1.23
1.45
C4A–C5
1.38
N1–C10A 1.39
N1–C11
N1–C2
C2–O2
N3–C2
N3–C31
N3–C4
C4–O4
C4–C4A
1.47
1.39
1.22
1.40
1.47
1.40
1.23
1.47
both in the ground and first excited singlet states. Dipole moments
in triplet states (T1) of both compounds are higher than in the other
states (l = 5.9 D for 138Me-5DAll and 5.8 D for 139Me-5DAll).
Table 9 collects the atomic charges of neutral forms of 138Me-
5DAll and 139Me-5DAll in the ground (S0) and in the first excited
singlet state (S1) in the gas phase, determined by the NBO popula-
tion analysis. Centres with partial negative charges include nitro-
gen atoms: N1, N3 and N10, oxygen atoms (O2 and O4) both in
S0 and S1 states (see also Fig. 3, for atom’s number). For both com-
pounds, the atoms with the largest positive charge are the carbonyl
atoms (C2 and C4), however the charges are larger at C2 than at C4,
and these charges do not change much in first excited singlet state.
Interestingly, we found some changes between electron density at
nitrogen atoms in the ground and in the first excited singlet states.
This trend is the same for both compounds, namely the magnitude
of the negative charges at nitrogen atoms decrease as follows:
N3 > N10 ꢁ N1 in the ground state and N3 ꢁ N10 > N1 in the
excited state (see also Table 9). Additionally the largest changes
in nitrogen atoms charges upon excitation occur at N10 nitrogen
atoms and these changes are similar in value for both, 138Me-
5DAll and 139Me-5DAll. In our previous work, we found that for
8-methyl-5-deazaalloxazine and 9-methyl-5-deazaalloxazine, the
magnitudes of the negative charge at nitrogen atoms in the ground
state are as follows, from lower to higher: N(3) > N(1) > N(10) [18].
As we found previously, that the largest growth of the negative
charge at heteroatoms of 8-methyl-5-deazaalloxazine and 9-
methyl-5-deazaalloxazine upon excitation occurs also on the N10
nitrogen atom of [18].
Fig. 6. The shape of the highest occupied molecular orbitals (HOMO) and the lowest
unoccupied molecular orbitals (LUMO) for 138Me-5DAll and 139Me-5DAll mainly
involved in the lowest singlet–singlet transitions. The isosurfaces correspond to the
wave function value of 0.02.
of 138Me-5DAll and 139Me-5DAll. We found that the S0 ? S1 tran-
sitions of both compounds have dominant contributions (66%)
from the HOMO ? LUMO excitations and are assigned as allowed
p,
p⁄ transitions.
In Table 7 we present results from the triplet state calculations;
the lowest T1 ? Ti excitation energies and oscillator strengths for
138Me-5DAll and 139Me-5DAll. The lowest excited triplet states
has the
p,
p⁄ character in both compounds. The predicted UV/Vis
transitions for 138Me-5DAll are located at 31,400, 25,800,
22,700, and 18,300 cmꢂ1. For 139Me-5DAll, these transitions are
situated at 31,900, 25,600, 22,800, and 19,000 cmꢂ1
.
Table
8 collects the calculated bond lengths and dipole
moments in the S0, S1 and T1 states for 138Me-5DAll and 139Me-
5DAll. Both compounds studied are planar in all the ground, first
excited singlet, and in the triplet states. The compounds are rela-
tively rigid, and the structural changes upon excitation are rela-
tively small. 139Me-5DAll has practically the same dipole
moments in its S0 and S1 states (
similar situation is for 138Me-5DAll (
tively), however 138Me-5DAll is more polar than 139Me-5DAll,
l
= 3.5 D and 3.3 D, respectively),
l
= 4.2 D and 4.1 D, respec-

146
D. Prukała et al. / Journal of Molecular Structure 1079 (2015) 139–146
Table 9
Atomic charges of neutral form of 138Me-5DAll and 139Me-5DAll in the ground (S0) and in the first excited singlet state (S1) in the gas phase determined by the NBO population
analysis.
Atom
138Me-5DAll
139Me-5DAll
Atom
138Me-5DAll
139Me-5DAll
S0
S1
S0
S1
S0
S1
S0
S1
C2
C4
C4A
C5
C5A
C6
C7
C8
C9
C81
0.853
0.688
0.851
0.684
0.853
0.699
0.852
0.694
C91
C9A
C10A
C11
C31
N10
N1
N3
O2
O4
–
–
ꢂ0.739
0.191
0.440
ꢂ0.743
0.192
0.443
0.199
0.452
0.202
0.456
ꢂ0.234
ꢂ0.082
ꢂ0.134
ꢂ0.202
ꢂ0.252
ꢂ0.013
ꢂ0.245
ꢂ0.743
ꢂ0.249
ꢂ0.061
ꢂ0.149
ꢂ0.195
ꢂ0.257
ꢂ0.014
ꢂ0.243
ꢂ0.745
ꢂ0.226
ꢂ0.091
ꢂ0.120
ꢂ0.217
ꢂ0.252
ꢂ0.229
ꢂ0.021
–
ꢂ0.242
ꢂ0.068
ꢂ0.135
ꢂ0.211
ꢂ0.257
ꢂ0.229
ꢂ0.016
–
ꢂ0.516
ꢂ0.519
ꢂ0.488
ꢂ0.481
ꢂ0.511
ꢂ0.630
ꢂ0.606
ꢂ0.516
ꢂ0.518
ꢂ0.499
ꢂ0.476
ꢂ0.508
ꢂ0.631
ꢂ0.596
ꢂ0.515
ꢂ0.519
ꢂ0.486
ꢂ0.478
ꢂ0.513
ꢂ0.629
ꢂ0.606
ꢂ0.515
ꢂ0.520
ꢂ0.496
ꢂ0.476
ꢂ0.509
ꢂ0.631
ꢂ0.597
[6] Y. Ikeuchi, M. Sumiya, T. Kawamoto, N. Akimoto, Y. Mikata, M. Kishigami, S.
Yano, T. Sasaki, F. Yoneda, Bioorg. Med. Chem. 8 (2000) 2027.
[7] Y. Kanaoka, Y. Ikeuchi, T. Kawamoto, K. Bessho, N. Akimoto, Y. Mikata, M.
Nishida, S. Yano, T. Sasaki, F. Yoneda, Bioorg. Med. Chem. 6 (1998) 301.
[8] J.M. Wilson, G. Henderson, F. Black, A. Sutherland, R.L. Ludwig, K.H. Vousden,
D.J. Robins, Bioorg. Med. Chem. 15 (2007) 77.
[9] H.I. Ali, N. Ashida, T. Nagamatsu, Bioorg. Med. Chem. 16 (2008) 922.
[10] J.H. Burckhalter, F.H. Tendick, E.M. Jones, P.A. Jones, W.F. Holcomb, A.L.
Rawlins, J. Am. Chem. Soc. 70 (1948) 1363.
[11] N. Keepyung, K.S. Won, Y.M. Kwang, Bioorg. Med. Chem. Lett. 3 (1993) 2631.
[12] J. Joule, G. Smith, Heterocyclic Chemistry, ELBS Low Price Edition, London,
1979. p. 126.
[13] S. Dudkin, V.O. Iaroshenko, V.Ya. Sosnovskikh, A.A. Tolmachev, A. Villinger, P.
Langer, Org. Biomol. Chem. 11 (2013) 5351.
Conclusions
We described spectral and photophysical properties of two tri-
methyl- derivatives of 5-deazaalloxazine using both experimental
and theoretical methods and compared them to those of their
alloxazine analogues. We conclude that although 5-deazaalloxa-
zines and alloxazines have similar structures, some differences in
their spectroscopy and photophysics are noted. The spectra of
138Me-5DAll and 139Me-5DAll, generally show a blue shift in
absorption and fluorescence maxima in comparison to their ‘‘aza’’
analogues.
[14] Z. Wang, C. Rizzo, J. Org. Lett. 2 (2000) 227.
Additionally, we found that solvent effects have small influence
on absorption and emission maxima of investigated compounds,
however solvents affects their fluorescence quantum yields and
fluorescence lifetimes. The values of fluorescence quantum yields
are higher and fluorescence lifetimes are longer in alcohols as com-
pared to those in aprotic solvents. According to theoretical calcula-
tions using TD-DFT methods for isolated molecules, we can state
[15] S. Dudkin, PhD thesis, Rostock, January 2013, Design and Synthesis of
Pyrimidine
C-Nucleosides,
5-Polyfluoroalkyl-5-deazaalloxazines
and
Spiro[pyrimido[4,5-b]quinoline-30,5-indoline-20-one]-3,10-dihydro-2,4-
diones.
[16] A. Koziołowa, V.N. Visser, J. Kozioł, M.M.M. Szafran, J. Photochem. Photobiol. A
93 (1996) 157.
[17] S. Salzmann, V. Martinez-Junza, B. Zorn, S.E. Braslavsky, M. Mansurova, C.M.
Marian, W. Gärtner, J. Phys. Chem. A 113 (2009) 9365.
[18] D. Prukała, M. Taczkowska, M. Gierszewski, T. Pe˛dzin´ ski, M. Sikorski, J.
that the lowest-energy transitions have p,
p⁄ character, being unac-
Fluoresc. 24 (2014) 505.
companied by the closely-located n,p⁄transitions, as is the case for
[19] E. Sikorska, I.V. Khmelinskii, W. Prukała, S.L. Williams, M. Patel, D.R. Worrall,
J.L. Bourdelande, J. Koput, M. Sikorski, J. Mol. Struct. 689 (2004) 121.
[20] E. Sikorska, I.V. Khmelinskii, W. Prukała, S.L. Williams, M. Patel, D.R. Worrall,
J.L. Bourdelande, J. Koput, M. Sikorski, J. Phys. Chem. A 108 (2004) 1501.
[21] M. Sikorski, D. Prukała, M. Insin´ ska-Rak, I. Khmelinskii, D.R. Worrall, S.L.
Williams, J. Hernando, J.L. Bourdelande, J. Koput, E. Sikorska, J. Photochem.
Photobiol. A 200 (2008) 148.
[22] S. Salzmann, Ch.M. Marian, Photochem. Photobiol. Sci. 8 (2009) 1655.
[23] E. Sikorska, I.V. Khmelinskii, J.L. Bourdelande, A. Bednarek, S.L. Williams, M.
Patel, D.R. Worrall, J. Koput, M. Sikorski, Chem. Phys. 301 (2004) 95.
[24] M. Bruszyn´ ska, E. Sikorska, A. Komasa, I. Khmelinskii, L.F.V. Ferreira, J.
Hernando, J. Karolczak, M. Kubicki, J.L. Bourdelande, M. Sikorski, Chem. Phys.
361 (2009) 83.
derivative of alloxazine. Also lowest triplet excited states of
138Me-5DAll and 139Me-5DAll have
p,
p⁄ character. On the basis
of theoretical calculations, we can also state, that the biggest
change in electron density, during S0 ? S1 transition, occurs at
nitrogen atom N(10), if we consider only the heteroatoms.
We described also the crystal structures of 138Me-5DAll and
139Me-5DAll, determined by X-ray diffraction methods. The crys-
tal structures of investigated compounds are similar, belonging to
the same space group and the molecules in these crystals are con-
nected by similar systems of hydrogen bonds.
[25] F. Yoneda, Y. Sakuma, J. Chem. Soc., Chem. Commun. 6 (1976) 2003.
[26] F. Yoneda, Y. Sakuma, S. Mizumoto, R. Ito, J. Chem. Soc., Perkin Trans. 1 (1976)
1805.
Acknowledgments
[27] H. Goldner, G. Dietz, E. Carstens, Liebigs Ann. Chem. 694 (1966) 142.
[28] ACD/ChemSketch; version 4.55, Advanced Chemistry, Development Inc.,
Toronto, Canada. Ed., 2000.
[29] Agilent Technologies, CrysAlis PRO (Version 1.171.33.36d), Agilent
Technologies, 2010.
[30] L.J. Farrugia, J. Appl. Cryst. 30 (1997) 565.
[31] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, J. Appl. Cryst. 26 (1993)
343.
This study was supported by the research grant DEC-2012/05/B/
ST4/01207 from The National Science Centre of Poland (NCN). All
of the calculations were performed at the PL-Grid project and by
´
the Poznan Supercomputing and Networking Centre.
[32] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[33] R. Ditcrfield, W.J. Herre, J.A. Pople, J. Chem. Phys. 54 (1971) 724.
[34] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
V.G. Zakrzewski, A.J. Jr. Montgomery, R.E. Stratmann, J.C. Burant, S. Dapprich,
J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V.
Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J.
Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, B.B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox,
T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, C.
Gonzalez, M. Head-Gordon, E.S. Replogle, J.A. Pople, GAUSSIAN 03, Revision
B05, Gaussian Inc, Wallingford, CT., 2003.
References
[1] L.V. Bystrykh, N.I. Govorukhina, L. Dijkhuizen, J.A. Duine, Eur. J. Biochem. 247
(1997) 280.
[2] D. Vargo, A. Pokora, S.-W. Wang, M.S. Johns, J. Biol. Chem. 256 (1981) 6027.
[3] P. Hemmerich, V. Massey, H. Fenner, FEBS Lett. 84 (1977) 21. and references
cited therein.
[4] A. Koziołowa, N.V. Visser, J. Kozioł, M.M. Szafran, J. Photochem. Photobiol. A 93
(1996) 157.
[5] T. Kawamoto, Y. Ikeuchi, J. Hiraki, Y. Eikyu, K. Shimizu, M. Tomishima, K.
Bessho, F. Yoneda, Y. Mikata, M. Nishida, K. Ikehara, T. Sasaki, Bioorg. Med.
Chem. Lett. 5 (1995) 2109.